The role of CREB3L4 in the proliferation of prostate cancer cells

The incidence of prostate cancer (PC) is growing rapidly throughout the world, in probable association with the adoption of western style diets. Thus, understanding the molecular pathways triggering the development of PC is crucial for both its prevention and treatment. Here, we investigated the role of the metabolism-associated protein, CREB3L4, in the proliferation of PC cells. CREB3L4 was upregulated by the synthetic androgen, R1881, in LNCaP PC cells (an androgen-dependent cell line). Knockdown of CREB3L4 resulted in decreased androgen-dependent PC cell growth. LNCaP cells transfected with siCREB3L4 underwent G2/M arrest, with upregulation of the proteins cyclin B1, phospho-CDK1, p21Waf1/Cip1, and INCA1, and downregulation of cyclin D1. Moreover, depletion of CREB3L4 resulted in significantly decreased expression of a subset of androgen-receptor (AR) target genes, including PSA, FKBP5, HPGD, KLK2, and KLK4. We also demonstrated that CREB3L4 directly interacts with the AR, and increases the binding of AR to androgen response elements (AREs). We also identified a role for the unfolded protein response (and its surrogate, IRE1α), in activating CREB3L4. Cumulatively, we postulate that CREB3L4 expression is mediated by an AR-IRE1α axis, but is also directly regulated by AR-to-ARE binding. Thus, our study demonstrates that CREB3L4 plays a key role in PC cell proliferation, which is promoted by both AR and IRE1α.

formation of abnormal epididymal sperm nuclei, and caspase activation 12 , leading to apoptosis of meiotic and postmeiotic germ cells 14 . CREB3L4 is abundantly expressed in the prostate, and is much more highly expressed in cancerous than non-cancerous prostate cells 15,16 . Specifically, CREB3L4 is abundantly expressed in high-grade prostatic intraepithelial neoplasia ( PIN), and all grades of adenocarcinomas, as compared to the normal prostate 15 . Thus, CREB3L4 could represent a potential biomarker for distinguishing benign from malignant prostatic cancer 17 . Although CREB3L4 expression is known to be regulated by androgen 15 , its role in prostate cancer progression and development is not well understood.
To this end, we explored a role for CREB3L4 in the androgen-dependent prostate cancer cell line, LNCaP. Our overall results demonstrate a role for CREB3L4 in modulating AR action, suggesting an interrelationship, and an AR-ER stress-CREB3L4 signaling axis, in prostate cancer cell proliferation.

CREB3L4 is overexpressed in androgen-dependent LNCaP cells, and is further induced by androgen.
To explore the biological role of CREB3L4 in prostate cancer, we first observed the expression level of CREB3L4 in prostate cancer cell lines, including the androgen-dependent prostate cancer (ADPC) cell line LNCaP, the androgen-insensitive metastatic subline C4-2B (derived from parental ADPC LNCaP cells), the androgen-independent prostate cancer (AIPC) cell line PC-3, and the nontumorigenic prostatic epithelial cell line RWPE-1. As shown in Fig. 1a, LNCaP cells expressed the highest levels of CREB3L4, fatty acid synthase ( FASN), and PSA, as compared to the other cell lines. Along with its mRNA, CREB3L4 protein expression was higher in LNCaP than other cell lines (Fig. 1b). Western blot revealed that CREB3L4 expression was significantly increased by the synthetic androgen R1881, compared to untreated LNCaP cells (Fig.1c). These results support previous data showing the expression of genes related to fatty acid and cholesterol biosynthesis, and IRE1α, all significantly increased by R1881, whereas the protein kinase RNA-like endoplasmic reticulum kinase (PERK)eukaryotic initiation factor 2 (eIF2α )-phosphorylation ER stress pathway, was suppressed by androgen 11,18 . From this background, we used LNCaP cells as an experimental model for studying androgen-dependent regulation of CREB3L4, with specific regard to prostate cancer cell proliferation.
CREB3L4 is required for LNCaP cell proliferation. We next investigated whether CREB3L4 affects the growth of LNCaP cells. For this, we transfected siCREB3L4, and observed proliferation of cells for a period of 8 days (D3, D6, D8) in the presence or absence of R1881 (Fig. 2a). LNCaP transfection with siCREB3L4 with or without R1881 inhibited proliferation at D3 by 20% (p < 0.05), D6 by 50% (p < 0.05), and D8 by 50% (p < 0.05), respectively. Specifically, from D3 to D8 after transfection, androgen-dependent proliferation of LNCaP cells was suppressed by siCREB3L4 transfection (Fig. 2a). siCREB3L4 also inhibited cell proliferation, even in the absence of androgen, at D8 (Fig. 2a). Microscopic observation also indicated that CREB3L4 knockdown resulted in LNCaP growth inhibition (Fig. 2b). We observed decreased numbers of G1 phase cells, with increased numbers of G2/M-arrested cells, regardless of the presence of androgen (Fig. 2c), although androgen even further increased G2/M arrest, by suppressing CREB3L4. These findings further support a role for CREB3L4 in LNCaP cell proliferation. To confirm whether CREB3L4 affects cell cycle regulation in cell growth, we assessed expression of proteins involved in the cell cycle, especially G2/M arrest-related proteins (Fig. 2d). We found that suppression of CREB3L4 resulted in the induction of cyclin B1, with repression of cyclin D1, causing mitotic clonal delay. Furthermore, expression of p21 Waf1/Cip1 and INCA1 (inhibitor of cyclin-dependent kinase (CDK) interacting with cyclin A1), in complex with CDK2, respectively 19 , was increased by siCREB3L4 (Fig. 2d). In addition, siCREB3L4 increased phospho-CDK1 with decreased cyclin D1 and CDK2 in the presence of androgen, making it likely that CREB3L4 affects G2/M arrest. These data suggest that CREB3L4 affects the proliferation of LNCaP cells by inhibiting cell cycle transitions.
CREB3L4 regulates AR-mediated transcription. Since AR plays a crucial role in prostate cancer cell proliferation, and is a common therapeutic target, we assessed possible roles for CREB3L4 in the AR signal cascade. Treatment of siCREB3L4-LNCaP cells with R1881 resulted in a significant reduction in protein levels of PSA, a well-known target of AR, even while AR expression was not changed (Fig. 3a). siCREB3L4 also decreased PSA expression levels in the absence of androgen (Fig. 3a). To confirm the effect(s) of CREB3L4 on the expression of various AR-downstream genes, their expression was observed in androgen-treated cells expressing siCREB3L4. Knockdown of CREB3L4 significantly decreased the expression of AR target genes, such as PSA, HPGD, FKBP5, KLK2, and KLK4, all of whose expression is tightly associated with prostate cell proliferation 20 (Fig. 3b). These data suggest that CREB3L4 may act as a positive regulator of AR.
CREB3L4 enhances AR activity through direct interaction with AR. Then, a question arises. What is the role of CREB3L4 in the AR-dependent signaling pathway? To examine the relationship between CREB3L4 and AR, with regard to prostate cancer cell proliferation, we performed reporter assays using a PSA gene promoter construct. Those reporter assays showed that human nuclear form of CREB3L4 (hCREB3L4N) activated the PSA promoter activity by 50-fold, in the absence of androgen and AR (Fig. 4a). This data is consistent with our other results showing that PSA expression is suppressed in siCREB3L4-LNCaP cells in the absence of androgen (Fig. 3a). These data indicate that PSA gene expression could be directly regulated by CREB3L4.
In the presence of androgen, however, AR activated the PSA gene promoter reporter by 25-fold (Fig. 4a). However, AR-mediated transactivation of PSA promoter activity in the presence of androgen was significantly increased, by 200-fold, when the hCREB3L4N was cotransfected (Fig. 4a). This result suggests that hCREB3L4N upregulates AR-mediated transactivation, in a ligand-dependent manner. Consequently, we next explored the mechanism of how CREB3L4 activates AR signaling. We first examined the interaction between CREB3L4 and AR by co-immunoprecipitation (co-IP), showing that the two proteins interact with each other (Fig. 4b).
Furthermore, the interaction between endogenous CREB3L4 and AR in LNCaP cells was also increased in the presence of androgen (Fig. 4c). To observe a possible role for CREB3L4 in AR-to-DNA recruitment, we performed chromatin immunoprecipitation (ChIP) assays. Androgen-induced AR recruitment to an androgen-response element (ARE), within the PSA gene promoter, was abolished in siCREB3L4-transfected cells (Fig. 4d). This result suggests that CREB3L4 interacts with AR, and assists AR recruitment to ARE. Taken together, these data suggest that not only can CREB3L4 upregulate the PSA gene AR-independently, it also enhances AR-mediated transactivation of its target genes by acting as an activator.

Androgen-induced CREB3L4 expression is regulated both directly, by AR, and indirectly, via
IRE1α pathway signaling. We next observed that androgen promotes LNCaP cell proliferation by inducing CREB3L4 expression, which cooperatively upregulates AR target genes. Thus, we speculated whether CREB3L4  in the presence or absence of 10-nM R1881 in 5% CT-FBS medium, were analyzed by propodium iodide (PI) staining and flow cytometry. As shown, there were significant differences in the G2/M population between scrambled versus siCREB3L4, in the absence/presence of R1881 (*P < 0.05), and siCREB3L4 versus siCREB3L4, in the presence of R1881 (#P < 0.05). (d) LNCaP cells were transfected with scrambled or siCREB3L4. After 24 hr, the cells were treated with R1881 (10 nM) or DMSO/EtOH (control) in 5% CT-FBS-containing medium, for 24-hr. Whole cell lysate protein (40-μ g) was resolved by SDS/PAGE, and immunoblotted for the respective proteins indicated, with GAPDH used as a loading control. Bars represent means ± S.E. (error bars), n = 3, *P < 0.05, scrambled versus siCREB3L4, in the absence/presence of R1881; #P < 0.05, scrambled versus scrambled, in the absence of R1881. DMSO, dimethyl sulfoxide; EtOH, ethyl alcohol; CT-FBS, charcoal treatedfetal bovine serum. may also play a critical role in regulating prostate cancer cell proliferation, through interacting with AR. We first questioned the molecular mechanism of how CREB3L4 gene expression is regulated in prostate cancer cells. It was reported that CREB3L4 expression is increased by androgen 15,16 , which we confirmed in Fig. 1c. Recently, canonical unfolded protein response (UPR) pathways were also shown to be directly and divergently regulated by androgens in prostate cancer cells, through modulation of AR transactivation 11 , including that of IRE1α , with inhibition of the PERK signaling pathway 11 . In addition, full-length CREB3L4 is cleaved to its nuclear form, upon ER stress, in LNCaP cells 21 . From this basis, we assumed that androgen-induced CREB3L4 expression could also be mediated through an AR-IRE1α signaling branch of the ER stress response. To test this hypothesis, we examined the relationship between IRE1α pathway signaling and CREB3L4 in AR-induced prostate cancer cell proliferation. These studies revealed that androgen-induced expression of CREB3L4 was inhibited by depletion of IRE1α (Fig. 5a), which also significantly decreased LNCaP cell proliferation. Ectopic expression of CREB3L4 in IRE1α-knockdown cells restored proliferation back to the level of control (i.e., non-transfected) LNCaP cells (Fig. 5b). Analogously, siCREB3L4-mediated inhibition of cell proliferation was further inhibited cell growth by siIRE1α (Fig. 5c). To examine the relationship between IRE1α and CREB3L4 in AR activity, we observed PSA expression levels following knockdown of IRE1α and/or CREB3L4. PSA levels were reduced by siCREB3L4, with even further reduction in cells cotransfected with siIREα (Fig. 5d). We next determined whether AR directly regulates CREB3L4 gene expression, via CHIP assays. Using Vector NTI Suite, we found four putative androgen response elements (AREs) (inverted repeated (IR) sequence-IR3; AGAACANNNTGTTCT; ARE1~ARE4) within the 3 kb region of the human CREB3L4 promoter (Supplementary Fig. 1a). Moreover, as might be expected, we observed that R1881 increases endogenous AR recruitment to the putative ARE4 region of the CREB3L4 promoter ( Supplementary Fig. 1b). These results suggest that androgen-induced CREB3L4 gene expression is in part regulated by AR directly. Furthermore, activation of ER stress response signals by AR contributes to CREB3L4 upregulation, resulting in activation of the androgen-dependent transcriptional cascades.

Discussion
CREB3L4 is highly expressed in the prostate gland, and upregulated even further in prostate cancer 15,22 . Indeed, CREB3L4 is now being evaluated as a potential marker for prostatic diseases, based especially on its ability to discriminate between benign and malignant prostate tumors 17 . However, the biological role of CREB3L4 in prostate cancer cell proliferation previously remained unknown. In this study, we found that CREB3L4 is essential  ) and/or AR, under the pGL3b or human PSA gene promoters, were transfected into HEK293 cells, followed by treatment with or without 10-nM R1881. After 24 hr, cell lysates were subjected to luciferase assay. Luciferase activities were normalized to Renilla luciferase activities to adjust for transfection efficiency. Normalized activities are shown as means ± S.E. (error bars), n = 3, and expressed as folds-increase, relative to basal activity. (b) HEK293 cells were transfected with AR, and CREB3L4 expression vectors, and proteins immunoprecipitated (IP' ed) using an anti-AR and anti-Flag for CREB3L4 antibodies. IP protein was then resolved by SDS/PAGE and immunoblotted for the respective proteins indicated. (c) Interaction of endogenous AR and CREB3L4 in LNCaP cells. Cells were treated with 10-nM R1881 for 24 hr. Endogenous AR and CREB3L4 protein from LNCaP cells were precipitated with anti-AR or anti-CREB3L4 antibodies, and the interaction between these proteins determined by co-IP. (d) ChIP assay performed in LNCaP cells transfected with siCREB3L4, with R1881 treatment. Normal IgG was used as a negative control for IP. AR response element (ARE) of PSA gene promoter was amplified to determine ChIP' ed DNA, which was normalized to total input DNA.
Scientific RepoRts | 7:45300 | DOI: 10.1038/srep45300 for prostate cancer cell proliferation, through its modulation of AR activity, which reciprocally, is upstream of CREB3L4 expression (perhaps suggesting some type of positive feedback). We also showed that CREB3L4 directly associates with AR, to enhance AR-induced transactivation of its downstream genes, in prostate cancer cells. Furthermore, we observed that AR-induced CREB3L4 expression is also mediated by IRE1α, ER stress response signaling. Our data thus suggest that the novel link between AR-CREB3L4 and the IRE1α pathway is an axis critical for driving prostate cancer progression (Fig. 5e).
The unfolded protein response (UPR) has critical roles in development and normal physiology, as well as in pathological states such as cancer 23 . The IRE1α -XBP1 pathway is known to be essential for tumor survival 24 , and loss of XBP1 sensitized cancer cells to death from oxidative stress 25 . Specifically, androgens induce a UPR response in prostate cancer cells, by activating the IRE1α -XBP1 signaling branch, to regulate growth and survival of prostate cancer cells 11 . XBP1 expression is also increased in clinical prostate cancer specimens 11 . Since CREB3L4 is upregulated by ER stress 14 , we assumed that androgen-induced CREB3L4 expression is mediated through the IRE1α pathway. We confirmed that AR-induced CREB3L4 expression is abolished through knockdown of IRE1α (Fig. 5a). In addition, we observed that CREB3L4 promoter activity is significantly increased by overexpressing hXBP1, a transcription factor gene downstream of the IRE1α pathway ( Supplementary Fig. 1c). These results suggest that androgen-induced CREB3L4 gene expression is in part regulated by AR directly (due to physical interaction with the CREB3L4 promoter), and in part mediated indirectly, via an IRE1α pathway. Thus, we posit that CREB3L4 is an essential mediator of AR-IRE1α -induced prostate cancer progression.
CREB3L4 knockdown also increased the number of cells in G2/M, and upregulated p21 Waf1/Cip1 and INCA1, both inhibitors CDK, regardless of the presence of androgen (Fig. 2c,d). However, phosphorylation of CDK1, after siCREB3L4 transfection, was further increased in the presence, but not absence, of androgen (Fig. 2d). Moreover, CDK1 phosphorylation coincided with a small, but statistically significant, G2/M arrest in siCRE-B3L4-transfected and R1881-treated cells, vs. androgen-untreated siCREB3L4 cells. This result shows that androgen could play an additive role in siCREB3L4-mediated arrest of G2/M phase prostate cancer cells. Indeed, we observed that siCREB3L4 transfection resulted in significant growth inhibition of LNCaP cells, in the presence of R1881, at D6 and D8 (Fig. 2a). This result suggests that androgen-induced CREB3L4 significantly increases cell proliferation.
There is a question whether CREB3L4 can even function in the absence of androgen. To address this, we showed that cell proliferation by D8, even in the absence of R1881, was decreased by siCREB3L4 (Fig. 2a,b). CREB3L4 also increased PSA expression independently of AR (Fig. 4a), and had a greater activating effect on the PSA promoter, compared to that of AR alone (Fig. 4a). Furthermore, transfection of siCREB3L4 decreased PSA protein levels in the absence of androgen (Fig. 3a). We also observed IRE1α downregulation, by CREB3L4 knockdown, in the absence of androgen (Fig. 5d). Thus, it is possible that IRE1α is also regulated by CREB3L4, similar to PSA. Therefore, we could not exclude androgen-independent effects of CREB3L4 on LNCaP cell proliferation. However, CREB3L4 had a more robust effect in the androgen-dependent state, compared to the androgen-independent state (Fig. 2a). We speculate that regulation of the AR-IRE1α -CREB3L4 pathway in the presence of androgen, may maximize the effect of androgen on proliferation of prostate cancer cells.
How does CREB3L4 regulate androgen-dependent cancer cell proliferation? We showed here that CREB3L4 knockdown resulted in decreased androgen-induced PSA expression, without affecting AR expression (Fig. 3a). Besides, CREB3L4 depletion decreased AR binding to an ARE within the PSA gene promoter (Fig. 4d). This indicates that CREB3L4 promotes AR binding to AREs, and transactivation of the associated target genes, by directly interacting with AR. However, we could not confirm recruitment of CREB3L4 to the ARE within the PSA gene promoter, nor binding to an unfolded protein response element (data not shown), in contrast to another study 26 . Thus, it yet remains necessary to identify CREB3L4-binding sites, within gene promoters, to better understand its regulation of PSA and other AR-target genes. We assume that independently bound CREB3L4 may interact with AR (bound to its ARE) on the PSA promoter, presumably by bending, "looping, " or some other unknown mechanism. Hence, more study is required to understand CREB3L4 activation of PSA, or other downstream genes, in a manner that is synergistic with AR.
Previous reports have demonstrated that the enzymes involved in fatty acid synthesis and cholesterol synthesis are upregulated in prostate cancer cells by androgens 27,28 . Thus, it was surprising to note that the expression of FASN was increased in siCREB3L4 knockdown cells. The mechanism of downregulation of FASN by the AR-CREB3L4 axis may be different from that of other AR target genes. Distinct cofactor complexes with CREB3L4/AR may occur in its regulation of lipogenesis-related genes. Thus, the molecular mechanism of this distinct regulation needs further study.
Collectively, we demonstrated that CREB3L4 is required for proliferation of prostate cancer cells, and that CREB3L4 is a crucial activator of AR function. Indeed, CREB3L4 directly interacts with, and facilitates, AR recruitment to the AREs of AR target genes, maximizing their expression, (e.g., PSA) (Fig. 3b). Additionally, we demonstrated that androgen-induced CREB3L4 expression is in part regulated by AR directly, and in part indirectly, by IRE1α signaling, suggesting that a distinct AR-ER stress-CREB3L4 regulatory axis also plays a role in prostate cancer proliferation. A schematic of the mechanism of action of CREB3L4, in facilitating prostate cancer proliferation, is shown in Fig. 5e. In summary, our findings demonstrate that CREB3L4 plays a key role in prostate cancer cell proliferation, justifying its further study as a possible prostate cancer biomarker and therapeutic target.

Methods
Cell culture and Transient Transfection Assay. HEK293 (embryonic kidney) and PC3 (androgen-independent prostate cancer cells) were obtained from ATCC (Manassas, VA, USA) and grown in Dulbecco's modified Eagle's medium (Hyclone, Logan, UT, USA), supplemented with 10% (v/v) fetal bovine serum, 100 units/ml penicillin, and 100-μ g/ml streptomycin. LNCaP (an androgen-dependent human prostate cancer cell line; ATCC number CRL1740) and RWPE-1 (nontumorigenic, prostatic epithelial cell line; ATCC number CRL11609) cells were grown at the Roswell Park Memorial Institute (RPMI, Buffalo, NY, USA) in medium 1640, supplemented with 10% FBS, and penicillin/streptomycin. C4-2B cell lines were supplied as a kind gift from Dr. H.G Yoon 29 . LNCaP cells were maintained in 10% FBS-containing media, and for all experiments, this media was replaced with 5% charcoal treated (CT)-FBS media to selectively remove hormones, growth factors, and steroids the day before observing the effects of R1881. Transient transfection and luciferase assays were performed using Fugene HD reagent (Promega, Madison, WI, USA) and a Dual luciferase assay kit (Promega). Luciferase activities were normalized to Renilla firefly activities, to adjust for transfection efficiency. Normalized luciferase activities are shown as means ± S.E. Data are expressed as fold-increases, relative to the basal activity of the reporters, in the absence of overexpression vectors.
Flow cytometry analysis. siRNA-transfected LNCaP cells were harvested by trypsinization, and fixed in ice-cold 70% ethanol overnight. The cells were then washed twice with cold PBS, and incubated for 30 min Scientific RepoRts | 7:45300 | DOI: 10.1038/srep45300 cells, using Lipofectamine RNA iMAX (Invitrogen) for at least 48 h, followed by cell lysis for RNA and protein preparation.